March 2, 2026 — BullshitBench v2 tests 70 models on hallucination, secretarial work disappeared across two decades, and Seedance 2.0 generates cinematic video from prompts
Favorites: Vintage Jets, Tax Brackets, and…
March 2, 2026 — BullshitBench v2 tests 70 models on hallucination, secretarial work disappeared across two decades, and Seedance 2.0 generates cinematic video from prompts